LP hosted its annual partner and client summit designed around holiday prep for eCommerce brands. The event, LP Connect, featured keynote sessions, sponsor-led panels, and a closing fireside chat with Portland Gear.
Below is a recap of the keynote topics and a full look at the panel discussions, including the insights brands and partners shared on stage.
Keynote Sessions
Four partners led the main stage, each tackling a different piece of the Q4 growth equation.
|
The Right Customer, Right Time: Winning with Deals & Offers (Presented by Tie) Opened the day with a look at how brands can use deals and offers strategically to reach the right customer at the right moment, rather than relying on blanket discounting. |
CTV & the Upper Funnel: Marketing Before the Click (Presented by MNTN) Covered how Connected TV fits into the upper funnel and what it takes to build demand before a customer ever reaches a site. |
|
After the Click: Tracking Calls, Forms & Growth Beyond Your Site (Presented by CallRail) Focused on the tracking gap that exists once a customer converts through a call or form, and how brands can close it to see the full picture of what’s driving growth. |
AI in Action: Smarter Attribution & Data Tracking (Presented by Triple Whale) Closed out the keynote lineup with a look at how AI is reshaping attribution and data tracking for eCommerce brands heading into Q4. |
Panel Discussions
Turning Existing Traffic into Lifetime Value
Presented by Klaviyo
Moderator: Ryan Garrow, LP Director of Partnerships and Client Acquisition
Panelists: Kendall McMullen (Klaviyo), Luke Schenacki (Ironman 4×4), Max Pond (FoamOrder)
The discussion centered on one theme: retention fuels growth when acquisition gets more expensive. Klaviyo advised consolidating email and SMS data into one system, since splitting the two channels across platforms leads to redundant messaging and misses the chance to personalize based on a full view of customer behavior. Its rule of thumb: any brand spending more than $10,000 a month on ads needs to treat email and SMS as the foundation the rest of the business runs on, not an expensive afterthought. FoamOrder and Ironman 4×4 built on that with real flow examples, using promotional discounts and timely reminders to move buyers past friction points in the funnel. Klaviyo also pointed to its customer agent tool as the connective layer across email, SMS, WhatsApp, and web chat, writing every interaction back into a single customer profile so follow-up messaging reflects the full conversation history, with Instagram DM support rolling out in Q4.
Key Takeaways:
- Watch for a declining email conversion rate year over year, even when revenue looks healthy. New customers can mask a retention problem.
- Use post-purchase surveys against product reviews to catch product issues before investing further in retention marketing.
- Test pop-up formats deliberately. FoamOrder’s best-performing version combined a side-entry format, the brand’s logo, and a callout of average order savings, beating generic messaging.
- Send campaigns to the most engaged segment first to improve inbox placement with ISPs like Gmail and Yahoo.
The First Impression: Landing Page Site Experience
Presented by Forix
Moderator: Karly Scott, LP Manager of Client Services
Panelists: Jesse Malti (Forix), Nhu-Chi Suarez (Tilly’s), Adam Gilbert (LP UX Designer)
The discussion centered on what happens once a click lands. An enriched product feed with accurate pricing, promotions, and inventory helps ad platforms like Meta match shoppers to the right signals before they ever reach a landing page, and the landing experience needs to continue the promise of the ad itself. Adam Gilbert introduced the law of single action: a landing page should make the next step obvious within about three seconds, since every additional decision is another chance for a shopper to leave. For brands without strong existing awareness, legitimacy signals, ratings, reviews, and localized inventory data, are the difference between a landing page that sells and one that stalls out.
As broad-match, AI-driven targeting options like Meta’s Advantage+ pull back some manual audience control, brands are responding by building more landing page variations, not fewer. Looking ahead, the panel expects AI-generated landing pages to push personalization further, dynamically building experiences per visitor based on segmentation data rather than relying on a fixed set of page variants.
Key Takeaways:
- A landing page needs to continue the promise of its ad. Shoppers coming from social convert in as few as zero to two clicks, and a mismatch shows up fast as high bounce rate, low scroll depth, and confused heat map movement.
- Keep the next step obvious and remove friction. Pop-ups that are hard to close on mobile were the most commonly overused element; a clear next step for shoppers not ready to convert, like related products, was the most underused.
- Personalization converts. Tilly’s stylist widget lifted conversion by about 2.5%, and if broad-audience traffic must land on a homepage, that page needs to look sharp and immediately legitimate.
- Test deliberately. Landing page tests can produce meaningful data in as little as two weeks, but watch for false positives from seasonality, promotions, overlapping tests, or a missing control group.
Cracking Amazon: The Next-Level Growth Channel
Presented by Threecolts
Moderator: Elisabeth Lageson, LP Enterprise Team Lead
Panelists: Heather Rodman (Weird Beverages), Craig Panozzo (LP Retail Media), Laura Koerber (Threecolts)
The sheer scale of the Amazon Marketplace set the stage: the platform captures 55% of online product searches and remains the largest online retailer by a wide margin over Walmart and every other marketplace combined. Brands that skip Amazon Advertising, or aren’t brand-registered, risk losing control of their own messaging to unauthorized sellers undercutting MAP pricing. Threecolts flagged the margin gap from FBA reimbursement leakage. Most brands are leaving 1% to 3% of FBA revenue on the table from missing or damaged inventory in shipments, which can add up to six figures for a $5 million to $10 million brand, and the same pattern is emerging on TikTok Shop as that platform’s fulfillment reporting matures. Looking forward, Amazon’s growing investment in agentic and AI-driven shopping, including new SKU-level profit analytics tools and MCP-based data access, signals a shift from keyword-based to more conversational, prompt-based targeting.
Weird Beverages shared their early experience, where the brand’s top-line numbers looked healthy, but hidden FBA storage fees and inefficient keyword spend quietly ate into margin. Building a model based on Amazon’s actual fee structure, then tracking it against monthly business reports, turned the channel into a self-funding profit center at scale.
Key Takeaways:
- Retail readiness comes first. Reviews, A+ content, and reliable fulfillment need to be in place before scaling ad spend, since ad spend can’t fix a listing that isn’t ready.
- Most brands leave 1% to 3% of FBA revenue on the table from missing or damaged inventory, which can add up to six figures for a $5 million to $10 million brand. The same leakage pattern is emerging on TikTok Shop.
- Test deliberately: title A/B tests (brand name first versus product descriptor first), A+ content and creative variations, and audience segments for both in-market shoppers and remarketing to past purchasers.
- Centralize financial reconciliation across marketplaces rather than reviewing Amazon, Walmart, Target, and D2C in isolation. It surfaces patterns, like a shared carrier issue, that a siloed view would miss.
AI Unfiltered — An Open Conversation
Presented by Feedonomics
Moderator: David Johnson, LP Chief Product Officer
Panelists: Kendall McMullen (Klaviyo), Jenna Gilardi (Feedonomics), John Tulien (LP), CP McBee (Microsoft)
The final panel packed a big punch with a plethora of insights on where AI shopping stands today. The conversation framed three parallel webs:
- Human web of search results, where most commerce still happens
- LLM web, where people shop through conversation
- Agentic web, where agents complete purchases on a shopper’s behalf
Brand consistency matters across all three, as clear signals help both human shoppers and AI agents interpret a brand the same way. Clarity, a free Microsoft tool, shows brands how their agentic traffic looks and how they’re showing up in AI conversations relative to competitors.
The current state of AI shopping resembles the early days of SEO: brands feeding LLMs the right attribute-level data— sizing, materials, fit, color, use case—are ranking easily right now, while product description pages alone don’t give AI systems enough to work with. AEO and GEO are ultimately an evolution of SEO, where accurate content and a clean product feed still matter, but more weight now falls on brand sentiment and third-party citations from sites like Reddit, LinkedIn, and YouTube. For brands with a limited budget: fix the product data foundation first, since a complete feed built for longer, more specific queries matters more now than one optimized for high-volume search terms.
Key Takeaways:
- Consumer trust in AI shopping is real. 60% of shoppers surveyed prefer an AI experience over a traditional website, and agentic traffic is growing eight times faster than human traffic.
- Feed LLMs attribute-level product data—sizing, materials, fit, color, use case—since product description pages alone aren’t enough for AI systems to work with.
- AEO and GEO are an evolution of SEO, not a replacement. The core work stays the same, but brand sentiment and third-party citations (Reddit, LinkedIn, YouTube) now carry more weight than the brand’s own site.
- Fix the product data foundation before investing in brand and creative. A complete feed built for longer, conversational queries matters more than one optimized for head terms.
- UCP (Universal Commerce Protocol) is emerging as the standard connecting agents and ad platforms, and product ads are currently the strongest-performing format in Copilot.
That’s a Wrap for LP Connect 2026
Special Guest: Portland Gear
The day closed with a fireside chat with Marcus Harvey, founder of Portland Gear, followed by a prize raffle and on-site happy hour. Marcus brought the energy home, running the crowd through a spontaneous rock, paper, scissors tournament before tracing Portland Gear’s path from a single Instagram account started in 2013 to a business on track for $30 million to $40 million in revenue this year, driven largely by its Cascade backpack line and a growing national footprint beyond Oregon. Looking ahead, he’s focused on finding a next hero product and watching AI-driven channels as a potential lower-cost path to new customers.
